Introduction

Red raspberry (Rubus idaeus L. var. Heritage) is one of the bramble cultivars with good flavor and attractive color. It grows in frigid and temperate regions, and its commercial planting regions mainly distributed in Poland, Jugoslavia, Russia, America, Canada, etc. At present, red raspberry fruits are widely harvested in northeast and northwest China.

Red raspberry is abundant in anthocyanins (Acys). As compared with other fruits, like miracle fruit (Acys content was 14.3 mg/100 g fresh weight), red raspberry is an excellent source of Acys [1]. It seems to be promising to use Acys extracts from red raspberry as a natural colorant and a natural antioxidant [2, 3]. Acys have a high free radical scavenging capacity, which have been proven by many reports. Joseph et al. [4] found that fruit extracts including Acys were effective in reversing age-related deficits in several neural and behavior degradation; Wagner [5] recognized that Acys were more effective than O-(β-hydroxyethyl) rutin in lowing the capillary permeability and fragility and in their anti-inflammatory and anti-oedema activities. The Acys in purple-colored sweet potato and red cabbage could suppress colon carcinogenesis induced by 1,2-dimethylhydrazine (DMH) and 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) in rats [3]. In addition, the most significant function of Acys is their ability to impart color to the plants or plant products, therefore, extraction of Acys from red raspberry is worth being deeply researched in the application to the functional natural colorant.

Usually, conventional solvent extraction of Acys is time and solvent consuming and has a low efficiency. Moreover, thermal extraction with a long time could cause the degradation of Acys and decrease the antioxidant activity of the extracts [6, 7]. Microwave-assisted extraction (MAE) utilizes the energy of microwaves to cause molecular movement and rotation of liquids with a permanent dipole leading to a very fast heating of the solvent and the sample, offering advantages like improved efficiency, reduced extraction time, low solvent consumption, and high level of automation compared to conventional extraction techniques [8, 9]. Recently, there have been many reports on the application of MAE on the extraction of natural products, such as artemisinin [10], ginseng saponins [11], and essential oils [12]. In addition, in MAE a wider range of solvents could be used, as the technique should be less dependent on a high solvent affinity [8].

Response surface methodology (RSM) is effective for responses that are influenced by many factors and their interactions, which was originally described by Box and Wilsonas [13]. Many studies indicated that it is useful for developing, improving, and optimizing processes [14, 15].

In this paper, the MAE parameters such as the ratio of solvents to materials, microwave irradiation power, and extraction time were optimized by RSM, in order to obtain the optimal extraction yield of Acys extracts from red raspberry. In addition, we identified the Acys composition of extracts by HPLC-MS and evaluated the influence of microwave on the extraction efficiency and chemical compositions of Acys extracts, in comparison to the conventional solvent extraction.

Materials and methods

Chemicals

Hydrochloric acid and 95% ethanol were of reagent grade, and methanol and formic acid were of chromatographic grade, and all chemicals were purchased from Beijing Chemical Reagents Company (Beijing, China). Amberlite CG-50 chromatographic ion exchange resin and a 0.45 μm membrane filter were purchased from Sigma Chemical Co.

Plant materials

Red raspberry fruits were provided by Forestry Institute of Chinese Academy (Beijing, China), and harvested from its orchards in July 2003. Each 1 kg of fruits were packaged in PE bag. After frozen at −40 °C for 48 h, all fruits were stored at −18 °C until further analysis.

MAE of anthocyanins from red raspberry

The MAE procedure that used in the experiment was developed by Kwon et al. [11] with a little modification. A total of 60±0.5 g of frozen fruits sample was crushed into pieces (3–4 mm) and put into a double-neck flask with a cooling system. After the flask was filled with the proper volume that was selected according to the solvent to sample ratio in experiments design, of extraction liquids, 1.5 M HCl – 95% ethanol (15:85), it was exposed for a period of time in a microwave extractor (Model NJL07-3, Jiequan microwave equipment Co., Ltd., Nanjing, China) under a given microwave power. The flask was taken out and cooled to room temperature by cooling water and filtered through Whatman No. 1 paper under vacuum, then collected in a volumetric flask. The residue was taken back and extracted again in the same conditions. The Acys extracts of the twice-extraction were mixed and used for determination of the total Acys content.

Conventional extraction of anthocyanins from red raspberry

The conventional extraction procedure used in this experiment has been described by Fuleki and Francis [16], and a little modification was made. A total of 60±0.5 g of frozen samples were smashed into pieces (3–4 mm), and mixed with 240 ml of 1.5N HCl – 95% ethanol (15:85) solvents in a 500 ml round bottom flask fitted with a cooling system. Extractions temperature was set at 55±1 °C, which was equivalent to mean temperature under optimized conditions in MAE process. Extractions were carried out for 12, 20, 30, 60 min, respectively. The Acys extracts was cooled to room temperature and filtered through Whatman No. 1 paper under vacuum and collected in a volumetric flask. The residue was taken back and extracted again in the same conditions. The Acys extracts of the twice-extraction were mixed and used for the determination of the total Acys content.

Purification of Acys extracts

Acys extracts of red raspberry were purified according to the procedure described by Fuleki and Francis [17]. The Amberlite CG-50 resin (50 g each time), was hydrated by placing in a beaker with repeated decantation and removed the fine particles with distilled water. Water slurry of the hydrated resin was poured into a 26×400 mm column, and the excess water was allowed to drain out without letting the column dry. Approximately 15 ml of aqueous extract was poured on the top of the column until the entire resin bed became red due to the absorbed Acys. Acys were absorbed onto the resin column while sugars, acids, and other water-soluble compounds were removed by washing the column with 100 ml of distilled water as judged by the refractive index of the liquid coming off the column. The pigments were eluted by adding ethanol containing 0.01% HCl (approximately 50 ml) until the resin returned to its original color. The acidified methanol fraction was concentrated with a rotary evaporator at 45 °C under vacuum until the ethanol was evaporated, and the residue was dissolved in 0.5% HCl solvent. The solution was stored at −20 °C until further analysis.

Determination of total anthocyanins content (T Acy) of red raspberry extract

T Acy was used to indicate the amount of Acys extracted from red raspberry extract. T Acy was determined with the method developed by Fuleki and Francis [17], and expressed as the amount of cyanidin-3-glucoside equivalents with unit of mg per 100 g fruit. The pigment contents were calculated according to the absorbance value at 533 nm of 1%/1 cm for Acys.

Identification of anthocyanins in red raspberry extracts by HPLC-MS

A 5 μl solution of the purified Acys samples was injected directly into HPLC-MS system (Agilent 1100 MSD series, Agilent, USA), which separated using a KROMASIL C18 (250 mm × 4.6 mm id, 5 μm) column fitted with a 20 mm × 4.6 mm id. KROMASIL C18 guard column (Eka Chemicals, Sweden) and detected at 520 nm. The mobile phase was composed of eluant A, formic acid – water, (3:97, v/v), and eluant B, formic acid – water – methanol (3:47:50, v/v/v). A gradient program was performed as follows: firstly, it was operated with the ratio of 60% A to 40% B for 2 min; and then within 40 min the solvent A was reduced to 0% and B was increased to 100%, at a flow rate of 0.8 ml/min.

Before being directed to an ion-trap electro-spray mass spectrometer equipped with an electro-spray ionization (ESI) and an ion-trap mass analyzer which was controlled by windows NT based Chem Station software (version 5.2, Agilent, USA) was used. The mass spectrometer was operated in the positive ion mode in the range 200–1000 m/z and under the following conditions: detector voltage, 3.5 kV; capillary voltage, 3500 V; capillary temperature, 250 °C; drying nitrogen gas at 350 °C was used at a flow rate of 10 l/min.

Tissue preparation and observation by scanning electron microscopy (SEM)

The fruit samples remained after conventional extraction and MAE respectively were fixed in 2.5% glutaraldehyde and 1% osmic acid for 12 h, and then dehydrated through an ethanol series in buffer: 30, 50, 70, 90, and 100% for 15 min each. Subsequently, the samples were dried in a critical-point dryer, and mounted on stubs, gold-coated then observed on the Hitachi S-5000 scanning electron microscope.

Statistical analysis

Statistical analysis was conducted with SPSS 10.0 software (version 10.0, SPSS Inc., USA). Trends were considered as significant when mean values of compared sets are different at P<0.05. SAS 9.0 software (version 9.0, SAS Institute Inc., USA) was used to design the CCRD and analyze the experiment result.

Results and discussion

Optimization of MAE condition of Acys extracts from raspberry by CCRD

The volume of solvent to be added would effect the extraction of the Acys due to possible evaporation losses or a non-complete interaction with the sample, and the microwave irradiation time and power applied effect the temperature obtained inside the vessels, which determines the efficiency of extraction [18]. In addition, results of the preliminary investigation of the MAE conditions has indicated that T Acys of red raspberry extracts were mainly dependent on the ratio of solvent to sample, the extraction time, and the irradiation power, therefore, a central composite rotate design (CCRD) was used to optimize the MAE conditions and investigate effects of three independent variables on T Acys of red raspberry extracts. The complete design consisted of 20 experimental points including six replications of the center points were shown in Table 1. The regression model was predicted by Eq. (1) as follows:

$$ Y = 16.3958 + 2.1327X_1 + 0.0642X_2 + 0.5364X_3\\ -\ 0.06134X_1^2 - 0.0018X_1 X_2 - 0.0001X_2^2 $$
(1)
Table 1 Experimental data and the observed response value with different combinations of extraction time (X 1), microwave irradiation power (X 2) and ratio of solvent to sample (X 3) used in the central composite rotate design (CCRD)
Full size table

The analysis of variance for these models is given in Table 2. According to this model, linear terms of extraction time (X 1, a<0.01), irradiation power (X 2, a<0.01), ratio of solvent to sample (X 3, a<0.05), quadratic terms of extraction time (\(X_1^2\), a<0.01), irradiation power (\(X_2^2\), a<0.01), and interaction of extraction time and irradiation power (X 1 X 2, a<0.01) reach significant. The result suggested that the change of the three factors, extraction time, irradiation power, and ratio of solvent to sample, had a significant effect on Acys content of extracts. In contrast, the interaction between extraction time and ratio of solvent to sample (X 1 X 3), irradiation power and ratio of solvent to sample (X 2 X 3), and quadratic term of ratio of solvent to sample (\(X_3^2\)) were not significant.

Table 2 Analysis of variance showed the effect of the processing variables as a linear term, quadratic term, and interactions on the responses considered
Full size table

Meanwhile, the whole model that include linear level (a<0.01), quadratic level (a<0.01), and cross-product level (a<0.05) all reached significant (a<0.01), but the lack of fit was not significant, which indicated an excellent adequacy of experiment value to predicted value. The optimal MAE conditions were obtained from response surface analysis as follows: extraction time: 12.1 min; irradiation power: 366 W, and solvent to sample ratio: 4/1. Under these conditions, the experimental T Acys (43.42 mg/100 g) was close to the predicted values (43.80 mg/100 g) calculated from the polynomial response surface model equation. Therefore, the RRCD was successful to predict the extraction efficiency of Acys from red cranberry by MAE.

In addition, when one factor was fixed as the optimal value calculated from above CCRD experiment, effects of another both factors on the extraction of Acys were shown by the contour optimizer plots. The effects of extraction time (X 1) and irradiation power (X 2) on the T Acys (Y) of extracts are reflected in Fig. 1. With the increase of X 1 and X 2, Y sharply mounted up, and then achieved saturated value when the extraction was conducted for 12 min at 360 W, and then did not rise anymore. The relationship between Y and irradiation power (X 2) as well as the ratio of solvent to sample (X 3) is illustrated in Fig. 2. The contour plot indicated that increases of X 2 and X 3 benefit to the extraction of Acys. However, the changes of X 2 have more significant effects on Y than those of X 3. As X 3 increased from 4/1 to 6/1, it hardly could make any promotion on Y. As for X 2, Y went up corresponsive with the increase of X 2, and reached the highest level at 360 W. Similarly, the effects of extraction time (X 1) and irradiation power (X 2) on T Acys (Y) of extracts are reflected in Fig. 3. X 3 is not as significant as X 1 in response for Y. Therefore, it could be concluded that the extraction time and irradiation power played prominent roles in getting high extraction efficiency of Acys during MAE; this was identical with the result reported by Eskilsson et al. [8] in study on the analytical-scale microwave-assisted extraction, due to acceleration of disruption of tissues and immigration of solutes from tissues under microwave power, which would be observed by SEM in this work, although the increase of solvent volume is always beneficial for the extraction of goal compounds [11].

Fig. 1
figure 1

Contour optimizer plot of total anthocyanins content (Y), extraction time (X 1), and irradiation power (X 2) (solvent to sample ratio was fixed at 4:1)

Full size image
Fig. 2
figure 2

Contour optimizer plot of total anthocyanins content (Y), irradiation power (X 2), and solvent to sample ratio (X 3) (extraction time was fixed at 10 min)

Full size image
Fig. 3
figure 3

Contour optimizer plot of total anthocyanins content (Y), extraction time (X 1), and solvent to sample ratio (X 3) (irradiation power was fixed at 300 W)

Full size image

Identifications of Acys extracts from red raspberry by HPLC-MS analysis

Gradient reversed-phase HPLC with absorbance detection and MS analysis were used to rapidly identify the main anthocyanins in red raspberry extracts. An A520 HPLC trace was illustrated in Fig. 4, mass spectral fragmentation patterns are shown in Fig. 5. For the standard Acys were unavailable, their identification was carried out by comparison of their retention time and by confirming the molecular weight with ES/MS, meanwhile previous studies of other researchers were taken as reference for us [19–22].

Fig. 4
figure 4

HPLC profile of anthocyanin extract by MAE. The peak of 13.3 min was cyanidin-3-sophoroside; the peak of 14.3 min was cyanidin-3-(2G-glucosylrutinoside); the peak of 15.8 min is cyanidin-3-sambubioside; the peak of 16.1 min is cyanidin-3-glucoside; the peak of 16.6 min was cyanidin-3-xylosylrutinoside; the peak of 17.4 min was cyanidin-3-(2G-glucosylrutinoside); the peak of 17.7 min was cyanidin-3-rutinoside; the peak of 21.5 min was pelargonidin-3-rutinoside; and the peak of 25.2, 28.2, 35.7, 39.2 min had not been given the special structures

Full size image
Fig. 5
figure 5figure 5figure 5figure 5

Positive ion mass spectra of anthocyanins composition in red raspberry extract: (1) cyanidin-3-sophoroside; (2) cyanidin-3-(2-glucosylrutinoside); (3) cyanidin-3-sambubioside; (4) cyanidin-3-glucoside; (5) cyanidin-3-xylosylrutinoside; (6) pelargonidin-3-(2-glucosylrutinoside); (7) cyanidin-3-rutinoside; (8) pelargonidin-3-rutinoside; (9), (10), (11), and (12) were not identified

Full size image

Identified results are shown in Table 3. There are 12 kinds of Acys in purified extracts. Seven kinds of major Acys included cyanidin-3-sophoroside, cyanidin-3-(2G-glucosylrutinoside), cyanidin-3-sambubioside cyanidin-3-rutinoside, cyanidin-3-xylosylrutinoside, cyanidin-3-(2G-glucosylrutinoside), and cyanidin-3-rutinoside. Other five kinds of Acys were minors, in which one Acys is identified as pelargonidin-3-rutinoside, but the other four kinds of anthocyanin failed to give their particular chemical structures, however, their aglycones were considered as pelargonin, petunidin, delphindin, malvidin, respectively.

Table 3 Identification of anthocyanin of red raspberry extracts by HPLC-MS
Full size table

Peak 1 (retention time (t R)=13.3 min), was cyanidin-3-sophoroside with molecular ion (M+) at m/z 610.9 and a fragment ion at m/z 286.7 which corresponds to cyanidin, and −324.2 was accorded with sophoroside.

Peak 2 (t R=14.3 min), with M+ at m/z 757.0, was identified as cyanidin-3-(2G-glucosylrutinoside). Fragment ion was at m/z 286.7, which corresponds to cyanidin, and m/z 610.8, with −146 as loss of rhamnose and −324 as loss for two molecules of indicans that was thought as rhamnoglucoside.

Peak 3 (t R=15.8 min), with M+ at m/z 580.8, was identified as cyanidin-3-sambubioside. The fragment ion at m/z 286.7 indicated it was cyanidin. Another fragment ion was at m/z 444.7, with −136.1 for xyloside, and −158 for another indican.

Peak 4 (t R=16.1 min), with M+ at m/z 448.7, was identified as cyanidin-3-glucoside. The fragment ion was at m/z 286.7, so it was cyanidin, with −162 corresponding to glucoside.

Peak 5 (t R=16.6 min), with M+ at m/z 726.8, was identified as cyanidin-3-xylosylrutinoside. The fragment ions were at m/z 286.7 for cyanidin, and m/z 448.7, with −162 for glucoside and −278.1 for two molecules of indicans.

Peak 6 (t R=17.4 min), with M+ at m/z 740.9, was identified as cyanidin-3-(2-glucosylrutinoside). The fragment ion at m/z 286.7 was for cyanidin. Another fragment ion was at m/z 594.8, with −308.1 for rutinoside and −454.2 for rutinoglucoside.

Peak 7 (t R=17.7 min), with M+ at m/z 594.8, was identified as cyanidin-3-rutinoside. The fragment ion was at m/z 286.7 for cyanidin, and −308 for loss of rutinoside.

Peak 8 (t R=21.5 min), with M+ at m/z 578.8, was identified as pelargonidin-3-rutinoside. The fragment ion was at m/z 270.7 for pelargonidin and −308 for rutinoside.

In addition, four kinds of Acys, i.e. peaks 9, 10, 11, and 12, were discovered in the red raspberry for the first time. The information of their molecular structure could be given as follows:

Peak 9 (t R=25.2 min) had M+ at m/z 596.8 and a fragment ion at m/z 330.7, corresponding to malvidin. A minor ion at m/z −266.1 which might be a molecule of pentose with a molecule of deoxygenated hexose.

Peak 10 (t R=28.2 min) had M+ at m/z 542.8 and a fragment ion at m/z 302.7 corresponded to delphindin. Another fragment ion was at m/z 432.6, with −240.2 for two molecules of pentose.

Peak 11 (t R=35.7 min) had M+ at m/z 470.7 and a fragment ion at m/z 316.7 accorded with petunidin, and the −154 fragment might be a molecule of hexose.

Peak 12 (t R=39.2 min) had M+ at m/z 753.1 and a fragment ion at m/z 267.5 might be pelargonidin. Another fragment ion was at m/z 398.5, with −132 fragment supposed to a molecule of pentose.

The major Acys components of red raspberry (Rubus idaeus L. var. Heritage) extract in our experiment were similar with the previous reports on red raspberry (Rubus idaeus L. var. Glen Ample) [20, 23, 24]. Meanwhile, some different kinds of Acys in red raspberry (Rubus idaeus L.), such as pelargonidin-3-sophoroside, pelargonidin-3-(2G-glucosylrutinoside), pelargonidin-3-glucoside, and cyanidin-3-sophoroside and cyanidin-3-(2G-glucosylrutinoside), which have been identified by Mullen et al. [19], have not been found in our experiment. The difference in Acys compositions probably correlated to varieties of red raspberry.

Comparison of MAE to conventional extraction of Acys in red respberry

Red raspberry samples were extracted by MAE and conventional extraction, respectively, in order to evaluate effects of MAE on the extraction efficiency and compositions of Acys. Results showed that MAE method was more efficient than the conventional extraction (in Table 4). When red raspberry samples were extracted for 12 min, T Acys by conventional method was only 70% of MAE. Even the conventional extraction took more time, such as 60 min, T Acys was just about 87% of MAE, although T Acys was increased as the extraction time prolonged. On other hand, the HPLC-MS analysis indicated that there was no significant difference in Acys compositions of extracts between MAE and conventional extraction (Fig. 6), therefore, MAE process did not result in any negative degradation reactions of Acys, this predicted MAE would not decline the bioactive function of red raspberry extracts.

Table 4 The comparison of the extraction efficiency by MAE method to conventional solvent extraction
Full size table
Fig. 6
figure 6

HPLC profile of anthocyanin extract by MAE and conventional extraction

Full size image

The microstructure of tissue of red raspberry after extraction was investigated by SME. The significant differences were revealed in Fig. 7A and B. The structure of red raspberry fruit after MAE was looser than that after conventional extraction. It seems that microwave irradiation results in an explosive disruption of the physical structure of fruit, leading to a direct migration of the desired Acys components into the surrounding solvent, because the selective and localized heating of residual moisture in fruits occurred under microwave irradiation. Similar results were investigated in soybean tissue by a high-intensity ultrasound-assisted extraction by Li et al. [25]. In conventional extraction process, a heated solvent slowly diffuses through the material, dissolving and carrying away target compounds. This direct mechanism means that MAE is a much more efficient and rapid extraction method.

Fig. 7
figure 7figure 7

Scanning electron micrographs of red raspberry fruit after extraction. A By conventional extraction. Very few cell walls were degraded by heating up. B By MAE. Microwave irradiation produced significant rupture of fruit tissues

Full size image

Conclusion

Microwave has the potential to be used in Acys extraction processes to improve efficiency and reduce processing time, and results obtained in this study would have implications for the functional natural Acys industry.

MAE conditions of Acys from red raspberry were optimized. When the MAE was conducted at the ratio of solvent to sample 4/1 for 12.1 min under 366 W, T Acys in extract would reach 98.33%, in which eight kinds of Acys were identified using HPLC-MS analysis, and cyanidin-3-sophoroside, cyanidin-3-(2-glucosylrutinoside), cyanidin-3-sambubioside, cyanidin-3-rutinoside, cyanidin-3-xylosylrutinoside, cyanidin-3-(2G-glucosylrutinoside), and cyanidin-3-rutinoside were seven kinds of main components. Our study demonstrated that a little difference in Acys composition of extract between MAE and conventional extraction, microwave irradiation did not destroy Acys in chemical structure.